“It feels like these data center companies have just put a big target on our backs,” said Kyle Schmidt, leader of the newly formed Protect Sand Springs Alliance. In Sand Springs, Oklahoma, the image that landed was not a rendering of sleek buildings or a promise of digital jobs. It was a parade float: a grim, smoking caricature of a data center looming over a gingerbread house. For a fast-growing class of industrial projects, the float captured an emerging reality. The toughest constraint on new data centers is no longer just chips, cooling, or power delivery; it is permission earned in public.
Across the United States, the same argument repeats with local variations: whether facilities designed to run around the clock, packed with servers and backup generators, belong on land that communities still describe as rural. In Sand Springs, leaders annexed 827 acres of agricultural property and began confidential talks with a large technology firm. Residents responded with crowded meetings, yard signs, and a newly formed group organized around blocking the project. The outrage has been sharpened by nondisclosure agreements, a common feature of corporate site selection that reads to many residents as a decision already made.
Data centers have always been physical infrastructure masquerading as something weightless. The artificial intelligence boom has ended that illusion. In the language of state regulators and grid planners, these buildings are “large loads.” In the language of neighbors, they are bright lights, noise, truck traffic, and an unfamiliar industrial footprint. The friction is not simply aesthetic. A single large data center can draw power on the scale of a midsize city, pulling new substations, transmission upgrades, and generation planning into places that may not have expected them. Those upgrades cost money, and the central question at local hearings is increasingly simple: who pays?
That question has moved beyond town halls and into the rulemaking machinery that governs electric rates. Three senators wrote that they were alarmed that residential electricity bills could “skyrocket,” and pointed to regions where increases reached as much as 267% over five years. At the same time, the nationwide average household electricity bill rose 7% year-over-year as of September, complicating efforts to isolate a single cause. The engineering detail underneath the politics is that today’s grid pricing model was not built for a world where a handful of new customers can arrive with demand measured in hundreds of megawatts, and then insist on custom contracts that may be shielded from public view.
In practical terms, communities are reacting to a design problem: the mismatch between local planning timelines and hyperscale development timelines. Developers move quickly once land and power look feasible. Municipalities, by contrast, must fit these projects into zoning codes, water plans, noise ordinances, and emergency response capacity. Even the definition of “industrial” becomes contested when a facility employs relatively few permanent workers but requires enormous continuous power and water support. In Menomonie, Wisconsin, a proposal billed as a $1.6 billion build on 320 acres promised up to 1,000 construction jobs and 50 to 75 full-time roles, while estimating about 75,000 gallons of water per day. The city’s biggest debate quickly became less about whether the site could be rezoned and more about whether residents were being asked to accept irreversible land-use change without basic information starting with who the ultimate operator would be.
Industry groups argue that opponents are misinformed and that communities are walking away from jobs and security. Yet the resistance is widening because the perceived risks are concrete and locally legible: rate impacts, water use, light pollution, and the secondary effects of building out power infrastructure. In multiple regions, the grid itself is signaling strain. Utility investment in transmission and distribution reached $84.9 billion in 2025, reflecting how quickly planners are trying to catch up to new load growth. The technical story interconnection queues, transformer shortages, congestion, and delayed upgrades has become, for many residents, a pocketbook story.
The most destabilizing element, however, may be procedural rather than electrical. Nondisclosure agreements, fast-tracked annexations, and confidential utility contracts have turned ordinary permitting fights into trust fights. “It’s the secrecy. The secrecy just drives people crazy,” said Jonathan Thornton, a realtor opposing a proposed project in Minnesota, in a similar dispute described in national reporting. When residents believe the process is opaque, even legitimate engineering claims about efficiency, conservation, or cost allocation tend to land as public relations.
Tracking groups have begun quantifying the consequences of this breakdown. Data Center Watch, run by 10a Labs, counted 20 projects blocked or delayed in a single quarter, a disruption totaling $98 billion in planned investment. In Sand Springs, officials emphasize the potential tax base and the appeal of a major employer, with the city manager noting the incentive of surpassing Walmart as the community’s largest taxpayer. The developer, White Rose Partners, has said that electricity-related costs would not fall on residential ratepayers and that the project would generate millions for schools and services.
Yet at the edge of a proposed campus, the argument compresses into plain language. “I don’t care how much chocolate icing you put on a dog turd, it don’t make it chocolate cake,” said Rick Plummer, a rancher who raises team-roping horses next to the planned site.
For data center developers, the engineering playbook is changing. Reliability and scale remain nonnegotiable, but the new baseline includes civic design: transparent project identities, intelligible resource accounting, and rate structures that withstand scrutiny. For communities, the debate has become a test of whether local governance can keep pace with infrastructure whose benefits are dispersed, while its externalities are experienced at the fence line. The outcomes increasingly hinge on something that cannot be procured in a supply chain: consent built early, in daylight, before the concrete arrives.
